A laboratory for tools that preserve knowledge and coordinate people.
We build AI tools that keep people, and the texture of human experience, at the center, never as something to extract from. It started as a thesis. Two are now in the world.
Four tools, one wager: that AI can be built for people at human scale, not to quietly extract from them, and that the alternative is something you can actually use.
Structural code intelligence.
Seshat compiles a codebase into a typed symbol graph, every function, route, and table with its real dependency edges, data flow, and architectural layer. Ask what breaks before you change it. Audit for layer violations, auth gaps, and dead code. Answers are exact, not guessed.
An operations spine for dog-walking businesses.
Multi-tenant software for the people who walk dogs for a living, and a community for the people whose dogs they walk. Recurring schedules, a real pricing engine, billing, group ventures, and meet-and-greets, the whole back office of a service business in one place.
Personal AI with emotional context. Multi-dimensional sentiment to help a person navigate their own record, and preserve the memories that carry weight.
Marketing attribution and ROI for small businesses. QR-tracked assets that show a local owner what their advertising actually does, in plain numbers.
Architectures of Empathy
We evolved to perceive what is safe, not what is real.
Ritual expanded what counted as safe. Each named deity was a container that made a domain of reality safe enough to perceive: safe enough to plant in, to fight in, to grieve in, to love in. A pantheon was never a collection of superstitions. It was a perceptual expansion toolkit, every god adding a dimension the body could finally attend to.
You can usually feel when something is built to take from you. Social media has that texture: the small extractions, the sense of being slowly mined. A lot of AI carries the same charge, and most of us register it long before we have words for it. That feeling is information. It is the body noticing an architecture that was not built for it.
So we build the opposite. An architecture of empathy keeps the person at the center, as the thing the system is for, held at the scale we actually evolved to trust: the group, the table, the people in the room. It is why these tools carry the old names. Seshat kept the records; her tool reads the structure of what we make. The name is the reminder of who the work is for.
AI does not have to be extractive. Built with care, it can draw us closer instead of mining us. The thesis has left the page; the rest of the arc is under construction.
Beyond the lab, I sit down with people one-on-one and work on their actual setup, your environment, your config, your workflow. You tell me what you are trying to do; I make it work, and you leave knowing how to do it yourself.
See how it works